Superoxide anion production is necessary for leukocyte, vascular endothelial and other functions during infection. However, excessive production of superoxide and its reactant products has been implicated in the pathogenesis of tissue injury and organ dysfunction occurring during sepsis and septic shock. Examination of tissue samples in animal models has also suggested that depletion of endogenous antioxidants such as superoxide dismutase during sepsis may potentiate this injury. As a result, antioxidant treatments employing low molecular weight nonprotein membrane-permeable superoxide dismutase mimetics have been developed for use in sepsis and other conditions associated with increased systemic inflammation. These agents which are metal-chelated macrocyclic ligand complexes, demonstrate free radical scavenging activities similar to superoxide dismutase. M40401 and M40403 are two such agents which show novel selectivity for superoxide anion itself. Superoxide anion however may have divergent effects during sepsis. In addition to the opposing roles superoxide might have on microbial killing and secondary inflammatory tissue injury, it also has the potential to alter hemodynamic function in opposing ways. Excessive superoxide production has been implicated in the oxidation of both endogenous and exogenous catecholamines which normally cause vasoconstriction. On the other hand, superoxide anion contributes to the inactivation of nitric oxide, a potent vasodilator. As a result, the overall effects of superoxide anion on vascular tone may represent its relative contribution to the inactivation of catecholamines versus nitric oxide. These contributions may vary during sepsis dependent on its severity. In turn, the hemodynamics effects of superoxide inhibitors like M40401 may also be influenced by the underlying severity of sepsis. The present studies investigated whether the severity of infectious challenge and its associated risk of death would alter the efficacy of M40401 in a rat model of sepsis. In individual experiments, animals were randomized to be challenged with doses of intravenous E. coli designed to produce low or high control mortality rates, following which they were treated with M40401 or placebo. The results showed that the efficacy of M40401 was dependent on control mortality rates. In experiments with high control mortality rates (i.e. > median), M40401 increased survival rates and mean arterial blood pressure and decreased platelet counts. However in experiments with low control mortality rates (i.e.